# Exploratory Cohort Study of Depressive Symptoms in South Africans with HIV-1 Subtype C: Associations with Kynurenine Pathway Metabolites and Inflammatory Markers

## TL;DR

This study explores how immune and metabolic factors in South Africans with HIV-1 subtype C may be linked to depressive symptoms and self-harm thoughts.

## Contribution

The study is the first to investigate kynurenine pathway metabolites and inflammatory markers in relation to depression in South Africans with HIV-1 subtype C.

## Key findings

- Higher baseline quinolinic acid (QUIN) was associated with increased depression risk and somatic symptoms.
- Lower baseline NGAL predicted greater odds of self-harm thoughts at follow-up.
- No baseline markers were associated with follow-up depression risk or scores.

## Abstract

Background: Depression is common in people living with HIV and may involve immune–metabolic dysregulation. Altered tryptophan (Trp) metabolism via the kynurenine (Kyn) pathway, particularly quinolinic acid (QUIN), has been linked to neuroinflammation and depression, yet data from South Africans with HIV-1 subtype C are limited. Methods: In this exploratory cohort, treatment-naïve adults with HIV were assessed at baseline in 2010 (n = 69) and followed up in 2015 (n = 40). Targeted LC-MS/MS quantified Trp–Kyn metabolites (Trp, Kyn, KA, QUIN) and ELISAs measured immune markers (suPAR, IL-6, hsCRP, sCD163, NGAL). Depressive symptoms were evaluated using the PHQ-9 at baseline and follow-up. Results: Baseline QUIN was associated with the PHQ-9 total score (adjusted R²=0.245; β=0.417; p = 0.004, p(Hochberg) = 0.044) and also increased odds of depression risk (OR=61.1; 95%CI 2.24–1664.76; p = 0.015, p(Hochberg) = 0.030) at baseline. Baseline NGAL was significantly lower in the follow-up group with self-harm thoughts compared to the group with no–self-harm thoughts (p = 0.008) and lower NGAL predicted greater odds of follow-up self-harm thoughts (OR = 0.007, 95% CI: <0.001–0.572, p = 0.027, p(Hochberg) = 0.027). When PHQ-9 subscales were examined, baseline QUIN, Kyn, hsCRP, and suPAR showed positive associations with the baseline somatic subscale (β range= 0.34–0.57, all p < 0.01), whereas the baseline cognitive–affective subscale was not associated with any marker. No baseline markers were associated with follow-up depression risk or scores. Over five years, suPAR decreased whereas hsCRP did not change significantly, however changes in these markers over time did not reveal any associations with the PHQ-9 total or subscale scores. Conclusions: In this exploratory cohort, higher peripheral QUIN may be linked to depression risk, particularly with somatic rather than cognitive–affective symptom dimensions. Lower baseline NGAL was the only marker associated with later self-harm thoughts, suggesting a potential neuroimmune signal that warrants further investigation. Larger, adequately powered longitudinal studies with repeated metabolomic assessments are warranted to confirm this finding and clarify temporal and causal relationships between immune–metabolic dysregulation and specific depressive symptom profiles.